How the recommendation search works

Most recommendation services only look at your top artists or recent plays. This one is different: it loads the complete listening history of your Last.fm profile — thousands of tracks across the years — and every track you have played becomes a starting point for the search. What you listened to, and how much, directly affects the weight of every recommendation.

For every track in your history the service fetches the Last.fm list of similar tracks and matches each candidate against everything you actually listen to. That adds up to tens of thousands of comparisons, so analyzing a large history can take tens of minutes. It is a deliberate trade-off: slower than usual recommendations, but far more accurate.

The result is a ranked list of new tracks that fit your musical profile best. All you need is a Last.fm username with scrobbles: start the search, leave the tab open, and come back to a finished list.